Prenatal Diagnosis of Congenital Toxoplasmosis with a Polymerase-Chain-Reaction Test on Amniotic Fluid

Patrick Hohlfeld, Fernand Daffos, Jean-Marc Costa, Philippe Thulliez, Francois Forestier, and Michel Vidaud ABSTRACT
Background Congenital infection with Toxoplasma gondii can produce serious sequelae. However, there is little consensus about screening during pregnancy, and the tests used to establish a prenatal diagnosis of toxoplasmosis are complex and slow. We evaluated a simpler approach that is based on a polymerase-chain-reaction (PCR) test. Methods Prenatal diagnostic tests, including ultrasonography, amniocentesis, and fetal-blood sampling, were performed in 2632 women with T. gondii infection acquired during pregnancy. In 339 consecutive women, a competitive PCR test for T. gondii was performed on amniotic fluid, and its results were compared with those of conventional diagnostic tests. The PCR test targets the B1 gene of T. gondii, uses an internal control, and can be completed in a day. Positive tests were confirmed by serologic testing of newborns or by autopsy in terminated pregnancies. Results Overall, the risk of fetal infection was 7.4 percent, but it increased sharply with gestational age. Congenital infection was demonstrated in 34 of 339 fetuses by conventional methods, and the PCR test was positive in all 34. In three other fetuses, only the PCR test gave positive results, and follow-up testing confirmed the presence of congenital toxoplasmosis. The PCR test gave one false negative result but no false positive results. The PCR test performed better than conventional parasitologic methods (sensitivity, 97.4 percent vs. 89.5 percent; negative predictive value, 99.7 percent vs. 98.7 percent). Conclusions For the prenatal diagnosis of congenital T. gondii infection, an approach based on a PCR test performed on amniotic fluid is rapid, safe, and accurate. Maternal infection with Toxoplasma gondii acquired during pregnancy may result in congenital infection and serious sequelae in the neonatal period or years after birth1. Prenatal diagnosis of congenital toxoplasmosis is based on ultrasonography,2 amniocentesis, and fetal-blood sampling3,4. Although reliable,5,6 screening pregnant women at risk and prenatal diagnosis of congenital toxoplasmosis remain

controversial7. Given the rate of false negative prenatal diagnoses,6 the risk associated with fetal-blood sampling, and the delay in obtaining definitive results with conventional parasitologic tests, better methods are needed. We evaluated a simpler approach to the prenatal diagnosis of toxoplasmosis. We report here on our experience with the conventional approach to the prenatal diagnosis of toxoplasmosis and compare it with the results of a test based on the polymerase chain reaction (PCR) and performed on amniotic fluid.

Methods
Study Population From 1983 to 1992, 2632 women with T. gondii infection acquired during pregnancy were referred to us. All prenatal diagnoses were performed in our unit. Before prenatal diagnosis, all patients started treatment with spiramycin (9 million IU [3 g] daily) as soon as their infection was confirmed or strongly suspected on the basis of the monthly serologic screening recommended in France. Prenatal Diagnosis Prenatal diagnosis was performed between 18 and 38 weeks of gestation by ultrasonography, amniocentesis, and fetal-blood sampling3. From September 1991 to December 1992, the results of a PCR test performed on amniotic fluid were studied prospectively and compared with the results of conventional tests in 339 consecutive women after informed consent was obtained. A fetus was considered to be infected if one or several of the specific diagnostic tests were positive -- that is, T. gondii were shown to be present in amniotic fluid or fetal blood by inoculation of mice,1 tissue culture,8 or the PCR test. Once the purity of the sample was confirmed,9 the presence of specific IgM in fetal blood as demonstrated by an immunosorbent agglutination assay was also considered a sign of infection when associated with positive parasitologic tests. Nonspecific biologic tests were used as aids in the decision-making process at the beginning of our project; they included the determination of total IgM levels, glutamyltransferase activity, and leukocyte, eosinophil, and platelet counts. Prenatal diagnosis was confirmed by postnatal serologic follow-up testing or by autopsy if the pregnancy was terminated. PCR Test For each sample, 1.5 ml of an aliquot was processed fresh for DNA amplification. The contaminating red cells were eliminated with a selective lysis buffer (0.32 M sucrose, 10 mM TRIS-hydrochloric acid [pH 7.5], 5 mM magnesium chloride, and 1 percent Triton X-100). After centrifugation, the pellets were resuspended in 50 microl of heat-detergent extraction buffer (10 mM sodium hydroxide, 0.5 percent polysorbate 20 [Tween 20], and 0.5 percent nonionic detergent [Nonidet P-40]). They were then heated for 10

minutes in boiling water and centrifuged at 16,000 x g for 10 minutes. The PCR was performed on 10 microl of the supernatant. The target of amplification was the 35-fold repetitive B1 gene of T. gondii (Table 1). To avoid false negative results we developed a positive internal control, using M13mp18 DNA10. Composite primers (Table 1) were used to generate a M13mp18 fragment with the T. gondii sequence incorporated at the ends by PCR. This fragment was then amplified with the T. gondii primers only to ensure a homogenized product of 143 base pairs. Approximately five molecules of internal control were added to each amplification reaction. Because this control contains the same primer template sequences, it competes with the T. gondii gene for primer binding and amplification. Thus, a result was considered negative only if the T. gondii gene was not amplified but the control sequence was amplified. This control procedure allowed monitoring of the sensitivity of the PCR in each sample. View this table: [in this window] [in a new window] Table 1. Sequences of the Primers Used to Amplify the T. gondii B1 Gene by Competitive Amplification.

To avoid contamination by the amplification products, deoxyuridine triphosphate nucleotides were substituted for deoxythymidine triphosphate nucleotides in the amplification-reaction mixture11. The samples were amplified in duplicate, in 50 microl of a reaction mixture containing 2 mM magnesium chloride; 50 mM potassium chloride; 10 mM TRIS-hydrochloric acid (pH 8.3); 0.01 percent (wt/vol) gelatin; 0.2 mM deoxyadenosine triphosphate, deoxyguanosine triphosphate, and deoxycytidine triphosphate; 0.4 mM deoxyuridine triphosphate; 20 pmol each of the T. gondii primers TOXO B22 and TOXO B23 (Genset, Paris); 1 microl of internal control (approximately five molecules); 0.5 U of uracil-DNA-glycosylase (GIBCO BRL, Gaithersburg, Md.); and 1.25 U of Taq DNA polymerase (Perkin-Elmer Cetus, Norwalk, Conn.). The samples were initially incubated for two minutes at 50 C to allow the uracil-DNAglycosylase to destroy any amplified product containing deoxyuridine triphosphate that could have been carried over from previous reactions. This incubation was followed by denaturation for five minutes at 95 C before temperature cycling. Forty cycles were performed, each cycle consisting of denaturation for 1 minute at 94 C, annealing for 30 seconds at 60 C, and extension for 1 minute at 72 C in a 48-well DNA thermal cycler (Perkin-Elmer Cetus). After 40 cycles, the primer extension was continued for 10 minutes at 72 C and an equal volume of chloroform was then added to inactivate the uracil-DNA-glycosylase.

Each amplification run contained several negative controls (heat-detergent extraction buffer with and without uracil-DNA-glycosylase) and positive controls (internal positive control in heat-detergent extraction buffer). To prevent contamination, strict partitioning of the different technical steps was observed12. Detection of PCR Products The amplification products were analyzed after electrophoresis on an 8 percent acrylamide gel and staining with ethidium bromide (Figure 1). To confirm the results and to allow semiquantitation, a reverse DNA hybridization kit (ToxoDNAgnostic, Genset) was used. In brief, biotinylated amplification products were captured in two microtiter wells by two oligonucleotide probes (specific for the B1 gene and M13mp18) covalently linked to the polystyrene wells. The detection system in the test uses fluorescence and is based on the streptavidin-alkaline phosphatase conjugate, which binds to the biotin of the amplified products. Fluorescent signals were measured with a fluorimeter (MicroFLUOR, Dynatech, Saint-Cloud, France) and were considered positive if their magnitude was at least five times that of the signals in the blank control. For the semiquantitative PCR,13 the result was interpreted as the ratio between values for the B1 gene and those for fluorescence from the internal control (Figure 1). The ratio of the fluorescence with the toxoplasma probe to that with the internal-control probe was used to define three semiquantitative categories: a large parasite burden, when the ratio was more than 4; a moderate burden, when the ratio was equal to or less than 4 but greater than 2; and a small burden, when the ratio was equal to or less than 2 but greater than 0.4. The relation between the ratio and the parasite burden was verified by determining serial dilutions of tachyzoites (data not shown). Figure 1. Amplification Products in Amniotic Fluid Subjected to the PCR Test for Congenital Toxoplasmosis. The amplification products were analyzed after acrylamide-gel (8 percent) electrophoresis and ethidium bromide staining. The markers were determined with use of the HaeIII fragment of X174; the fragments 143 base pairs (bp) long correspond to the amplification of the internal control, and 115-bp fragments to the amplification of T. gondii B1 gene. Lane 1 represents a negative sample; lanes 2 through 4 represent positive samples; and lane 5 represents a sample without amplification. View larger version The lower panel shows the fluorescence ratios corresponding to (91K): the lanes above. The ratio of the fluorescence of the toxoplasma [in this window] probe to the fluorescence of the internal-control probe [in a new window] (TPF:ICPF) was obtained by measuring the fluorescence signals in two microtiter wells. The result was considered negative when the ratio was 0.2 or less (lane 1); positive when the ratio was 0.4

or more, provided that the toxoplasma-probe fluorescence signal was five times the blank-control signal (lanes 2, 3, and 4); and inconclusive, whatever the ratio, when both the toxoplasma-probe signal and the internal-control-probe signal were less than five times the blank-control signal (lane 5). The TPF:ICPF ratio was used for semiquantitation of positive samples: lane 2, 7.90 (large parasite burden); lane 3, 1.20 (small parasite burden); and lane 4, 3.19 (moderate parasite burden). Statistical Analysis The statistical analysis was mainly descriptive, and the 95 percent confidence intervals were calculated according to a binomial distribution.

Results
The time of maternal infection was reliably established in 2281 of the 2632 women on the basis of the monthly screening. The overall risk of fetal infection was 7.4 percent but varied with gestational age (Table 2). Although there were only 100 pregnancies in which the mothers became infected during the two weeks after the last menstrual period, it appears that such early infections posed little or no risk to the fetus. View this table: [in this window] [in a new window] Table 2. Incidence of Congenital Toxoplasmosis According to Gestational Age at the Time of Maternal Infection.

Of the 194 cases of congenital toxoplasmosis that we identified, 178 were diagnosed by conventional methods of prenatal diagnosis (i.e., inoculation of mice with amniotic fluid and fetal blood, tissue culture of amniotic fluid, and the identification of specific IgM in fetal blood). There were no false positive results. The overall sensitivity was 92 percent (95 percent confidence interval, 88 to 96 percent), the specificity 100 percent (95 percent confidence interval, 99.8 to 100 percent), and the negative predictive value 99 percent (95 percent confidence interval, 99.0 to 99.7 percent). The sensitivity of parasitologic tests ranged from 64 to 72 percent (Table 3). The sensitivity of specific IgM increases with gestational age: the rate was 12 percent among infected fetuses in which blood

samples were obtained before the gestational age of 25 weeks, as compared with 39 percent between 25 and 30 weeks and 59 percent after 30 weeks. View this table: [in this window] [in a new window] Table 3. Sensitivity of the Conventional Tests Used for Prenatal Diagnosis of Congenital Toxoplasmosis.

An increased leukocyte count was found in 7 percent of infected fetuses, eosinophilia in 9 percent, thrombocytopenia in 27 percent, an increased level of total IgM in 37 percent, and increased -glutamyltransferase activity in 36 percent. None of these factors were predictive of the severity of fetal lesions. Termination of pregnancy was considered if ultrasonograms revealed major lesions in the fetus (mainly ventricular dilatation) or if infection of the fetus was confirmed after the mother had contracted toxoplasmosis during the first 10 weeks of pregnancy. The pregnancy was terminated in 73 cases (2.8 percent of those referred), in all cases because of maternal infections during the first half of pregnancy (range, 3 to 22 weeks). No case of ventricular dilatation was identified among mothers with seroconversion after the 22nd week. The outcome of pregnancy was uneventful for most uninfected fetuses. Early fetal loss -- that is, within 3 weeks after sampling -- occurred in 16 cases (0.6 percent), and intrauterine death in 18 additional cases 27 to 132 days after the procedure (overall rate of spontaneous fetal loss, 1.3 percent). Prenatal diagnoses were reached in 339 consecutive cases with the use of both conventional methods and the PCR test. In 13 cases, there was no positive signal from the internal-control probe and the test had to be repeated with another aliquot. This problem was solved when a second assay was negative, and in no case was a second sample required. Congenital infection was demonstrated in 34 fetuses by conventional methods. The PCR test was positive in all 34 cases. In three additional cases, the PCR test was the only positive test, and congenital infection was ultimately confirmed by autopsy findings in two cases and by serologic follow-up testing of the infant in one case. In this series, there were no false positive results of prenatal diagnosis. All positive tests correlated with active fetal disease. There was one false negative result. Maternal infection occurred around the 18th week of pregnancy; when performed 4 weeks later, all diagnostic tests were negative. During the 34th week, ultrasonography revealed

intracranial densities and led to further sampling. The PCR test and inoculation of mice with fetal blood then gave positive results. The diagnostic value of the PCR test and conventional tests is shown in Table 4. The relative risk of congenital infection (positive predictive value/1 - negative predictive value) was 333 for the PCR test and 77 for the conventional methods. View this Table 4. Diagnostic Value of the PCR Test as Compared with table: Conventional Methods of Prenatal Diagnosis of Congenital [in this Toxoplasmosis in 339 Pregnancies. window] [in a new window]

The results of semiquantitative PCR testing with the B1 gene were verified with the use of a unique copy gene (P30). They showed that most positive samples had a moderate parasite burden, for which the sensitivity of tissue culture equaled that of inoculation of mice. When the results of semiquantitative PCR testing were compared with those of tissue culture and inoculation of mice (Table 5), it appeared that when the parasite burden was small, the sensitivity of inoculation of mice was poor and that of tissue culture was inefficient. In the three cases shown to be positive by only the PCR test at the time of prenatal diagnosis, the parasite burden was small. On the other hand, inoculation of large numbers of parasites does not always induce a detectable reaction in mice. No correlation was observed between the results of semiquantitation and fetal outcome. View this Table 5. Positivity of Amniotic Fluid in Tissue Culture and Inoculated table: Mice, According to Semiquantitative PCR Category of Parasite Burden in [in this 37 Pregnancies. window] [in a new window]

Discussion
The prenatal diagnosis of congenital toxoplasmosis is important to prevent unnecessary termination of pregnancy. After experience with more than 2500 cases, the risk of fetal infection can now be precisely described. Very early infection of the mother (within two weeks of the last menstrual period) poses little or no risk to the fetus, which is important since when first seen, most patients are at least two months pregnant. A positive

screening test (for IgG and IgM) could be interpreted as showing a very low risk if the titer of IgG remained stable in two specimens obtained three weeks apart1 and run in parallel, with the first sample obtained before the 10th week of pregnancy. In this situation, a consistently high IgG titer indicates that the infection was most probably acquired more than two months earlier1,14. Although it is impossible to conclude that the risk is absolutely zero, since cases of congenital toxoplasmosis due to maternal infection before pregnancy have been described,1,15,16,17 that possibility does not seem to warrant further prenatal diagnostic testing. Several studies have described the results obtained with the PCR in the diagnosis of toxoplasmosis with use of P3018,19 and B1 gene targets20,21 or a segment of the 18S ribosomal DNA22,23. In our series, the PCR test based on the B1-gene target in amniotic fluid had a higher sensitivity than conventional methods, and its results can be obtained within 24 hours. In this study, the absence of false positive results shows the efficacy of specific decontamination with uracil-DNA-glycosylase to prevent carryover contamination. In addition, the sensitivity of the PCR was continuously monitored for each sample at each assay with an internal competitive control. This is of utmost importance since DNA amplification may be the only positive result when the concentration of parasite is low in the sample. Competitive PCR testing allowing semiquantitation of parasites in samples shows the low sensitivity of conventional methods when the number of parasites is low and the possibility of an absence of reaction in mice when the number is very high, as described by others24. Thus, delayed transplacental transfer does not explain all false negative results of conventional prenatal diagnostic testing. The PCR produced one false negative result, and thus seems the most sensitive and reliable method at present. This false negative result was not due to an absence of the B1 gene in the parasite, since the test was positive 12 weeks later. The possibility of false negative results shows the necessity of regular serologic follow-up testing of all infants. In the past we used nonspecific biologic tests to evaluate the risk of fetal infection and determine the need for therapy, while awaiting the results of the parasitologic tests. Since such biologic tests do not have considerable prognostic value, we now rely solely on amniotic-fluid studies to assess fetal risk. The lack of sensitivity of tissue culture as compared with the PCR test led us to abandon the former. We propose that inoculation of mice with amniotic fluid be retained as a confirmatory method, given that the quality of testing with PCR may vary among laboratories. Moreover, inoculation of mice might allow the isolation and conservation of T. gondii strains25. The rate of spontaneous adverse outcomes of pregnancy was low in this study and was comparable to the risk posed by amniocentesis26,27,28. In most centers, however, fetalblood sampling entails a substantially greater risk of fetal loss than does

amniocentesis29,30. This new approach to sampling makes the prenatal diagnosis quicker and safer and thus more acceptable to patients. A reliable, safe, rapid, simple, and cheaper method of prenatal diagnosis by PCR is now available and can be used from the 18th week of pregnancy until term. This should be taken into account if a policy of screening during pregnancy is considered. Despite reasonable evidence of the efficacy of the treatment of infected fetuses,5,6,31,32,33,34 the pros and cons of obstetric screening for toxoplasmosis have been discussed for years7,16,32,35,36,37,38,39,40 and have been recently reviewed41. It seems logical to try to reduce the risk of infection by adequate health education,37 but the effectiveness of primary prevention through health-education programs is uncertain,1,42 and until now, screening together with prenatal diagnosis has remained the only possibility for a pregnant woman to reduce the risk of giving birth to an infant severely affected by toxoplasmosis. Screening without specific prenatal diagnosis carries the risk of unnecessary termination of pregnancy or a burden of anxiety throughout pregnancy and the infant's first months of life. Maternally transmitted IgG takes 8 to 12 months to disappear. Many infants would be exposed for months to a treatment that might have side effects, and only a small proportion of such infants would actually benefit from it. In this study, amniocentesis was always performed together with fetal-blood sampling to compare the performances of both methods. Our experience involved only prenatal diagnosis performed after the 18th week of pregnancy. The reliability of the PCR test before this point remains unknown. Finally, prenatal diagnosis should not be attempted until at least four weeks after acute disease in the mother. In conclusion, this study shows how the approach to prenatal diagnosis of congenital toxoplasmosis has evolved since our first descriptions3,5. Prenatal diagnosis is not warranted if maternal infection occurs during the first two weeks of pregnancy. Fetalblood sampling is no longer necessary. Amniocentesis together with the PCR test and inoculation of mice is a safer means to obtain reliable diagnostic information within one day of sampling.
Supported in part by a grant (9.1.90) from the Caisse Regionale d'Assurance Maladie d'Ile de France. We are indebted to Yvette Sole for technical assistance with the polymerase chain reaction and to Luc d'Auriol, Ph.D., and Thierry Poynard, M.D., for helpful comments on the manuscript.

Source Information
From the Service de Medecine et de Biologie Foetales (P.H., F.D., J.-M.C., F.F., M.V.) and Laboratoire de la Toxoplasmose (P.T.), Institut de Puericulture de Paris, Paris.

Address reprint requests to Dr. Vidaud at the Service de Medecine et de Biologie Foetales, 26, Blvd. Brune, 75014 Paris, France.

References
1. Remington JS, Desmonts G. Toxoplasmosis. In: Remington JS, Klein JO, eds. Infectious diseases of the fetus and newborn infant. 3rd ed. Philadelphia: W.B. Saunders, 1990:89-195. 2. Hohlfeld P, MacAleese J, Capella-Pavlovsky M, et al. Fetal toxoplasmosis: ultrasonographic signs. Ultrasound Obstet Gynecol 1991;1:241-244.[CrossRef] [Medline] 3. Desmonts G, Daffos F, Forestier F, Capella-Pavlovsky M, Thulliez P, Chartier M. Prenatal diagnosis of congenital toxoplasmosis. Lancet 1985;1:500-504. [Medline] 4. Daffos F, Capella-Pavlovsky M, Forestier F. Fetal blood sampling during pregnancy with use of a needle guided by ultrasound: a study of 606 consecutive cases. Am J Obstet Gynecol 1985;153:655-660.[Medline] 5. Daffos F, Forestier F, Capella-Pavlovsky M, et al. Prenatal management of 746 pregnancies at risk for congenital toxoplasmosis. N Engl J Med 1988;318:271275.[Abstract] 6. Hohlfeld P, Daffos F, Thulliez P, et al. Fetal toxoplasmosis: outcome of pregnancy and infant follow-up after in utero treatment. J Pediatr 1989;115:765769.[Medline] 7. McCabe R, Remington JS. Toxoplasmosis: the time has come. N Engl J Med 1988;318:313-315.[Medline] 8. Derouin F, Thulliez P, Candolfi E, Daffos F, Forestier F. Early prenatal diagnosis of congenital toxoplasmosis using amniotic fluid samples and tissue culture. Eur J Clin Microbiol Infect Dis 1988;7:423-425.[Medline] 9. Forestier F, Cox WL, Daffos F, Rainaut M. The assessment of fetal blood samples. Am J Obstet Gynecol 1988;158:1184-1188.[Medline] 10. Bretagne S, Costa JM, Vidaud M, Tran J, Nhieu V, Fleury-Feith J. Detection of Toxoplasma gondii by competitive DNA amplification of bronchoalveolar lavage samples. J Infect Dis 1993;168:1585-1588.[Medline]

11. Longo MC, Berninger MS, Hartley JL. Use of uracil DNA glycosylase to control carry-over contamination in polymerase chain reactions. Gene 1990;93:125-128.[CrossRef][Medline] 12. Kwok S, Higuchi R. Avoiding false positives with PCR. Nature 1989;339:237238. [Erratum, Nature 1989;339:490.][CrossRef][Medline] 13. Siebert PD, Larrick JW. PCR MIMICS: competitive DNA fragments for use as internal standards in quantitative PCR. Biotechniques 1993;14:244-249. [Medline] 14. Thulliez P, Daffos F, Forestier F. Diagnosis of Toxoplasma infection in the pregnant woman and the unborn child: current problems. Scand J Infect Dis Suppl 1992;84:18-22.[Medline] 15. Mitchell CD, Erlich SS, Mastrucci MT, Hutto SC, Parks WP, Scott GB. Congenital toxoplasmosis occurring in infants perinatally infected with human immunodeficiency virus 1. Pediatr Infect Dis J 1990;9:512-518.[Medline] 16. Desmonts G, Couvreur J, Thulliez P. Toxoplasmose congenitale: cinq cas de transmission a l'enfant d'une infection maternelle anterieure a la grossesse. Presse Med 1990;19:1445-1449.[Medline] 17. Garcia AG. Congenital toxoplasmosis in two successive sibs. Arch Dis Child 1968;43:705-710.[Medline] 18. Dupouy-Camet J, Bougnoux ME, Lavareda de Souza S, et al. Comparative value of polymerase chain reaction and conventional biological tests for the prenatal diagnosis of congenital toxoplasmosis. Ann Biol Clin (Paris) 1992;50:315-319. [Medline] 19. Savva D, Morris JC, Johnson JD, Holliman RE. Polymerase chain reaction for detection of Toxoplasma gondii. J Med Microbiol 1990;32:25-31.[Abstract] 20. Grover CM, Thulliez P, Remington JS, Boothroyd JC. Rapid prenatal diagnosis of congenital Toxoplasma infection by using polymerase chain reaction and amniotic fluid. J Clin Microbiol 1990;28:2297-2301.[Medline] 21. van de Ven E, Melchers W, Galama J, Camps W, Meuwissen J. Identification of Toxoplasma gondii infections by B1 gene amplification. J Clin Microbiol 1991;29:2120-2124.[Medline]

22. Cazenave J, Forestier F, Bessieres MH, Broussin B, Begueret J. Contribution of a new PCR assay to the prenatal diagnosis of congenital toxoplasmosis. Prenat Diagn 1992;12:119-127.[Medline] 23. Guay JM, Dubois D, Morency MJ, Gagnon S, Mercier J, Levesque RC. Detection of the pathogenic parasite Toxoplasma gondii by specific amplification of ribosomal sequences using comultiplex polymerase chain reaction. J Clin Microbiol 1993;31:203-207.[Abstract] 24. Derouin F, Mazeron MC, Garin YJF. Comparative study of tissue culture and mouse inoculation methods for demonstration of Toxoplasma gondii. J Clin Microbiol 1987;25:1597-1600.[Medline] 25. Sibley LD, Boothroyd JC. Virulent strains of Toxoplasma gondii comprise a single clonal lineage. Nature 1992;359:82-85.[CrossRef][Medline] 26. Tabor A, Philip J, Madsen M, Bang J, Obel EB, Norgaard-Pedersen B. Randomised controlled trial of genetic amniocentesis in 4606 low-risk women. Lancet 1986;1:1287-1293.[CrossRef][Medline] 27. Smidt-Jensen S, Permin M, Philip J, et al. Randomised comparison of amniocentesis and transabdominal and transcervical chorionic villus sampling. Lancet 1992;340:1237-1244.[CrossRef][Medline] 28. Halliday JL, Lumley J, Sheffield LJ, Robinson HP, Renou P, Carlin JB. Importance of complete follow-up of spontaneous fetal loss after amniocentesis and chorion villi sampling. Lancet 1992;340:886-890. [Erratum, Lancet 1992;340:1236.][CrossRef][Medline] 29. D'Alton ME, DeCherney AH. Prenatal diagnosis. N Engl J Med 1993;328:114120.[Full Text] 30. Ghidini A, Sepulveda W, Lockwood CJ, Romero R. Complications of fetal blood sampling. Am J Obstet Gynecol 1993;168:1339-1344.[Medline] 31. Couvreur J, Thulliez P, Daffos F, et al. Foetopathie toxoplasmique: traitement in utero par l'association pyrimethamine-sulfamides. Arch Fr Pediatr 1991;48:397403.[Medline] 32. Aspock H, Flamm H, Pircher O. Die Toxoplasmose-Uberwachung wahrend der Schwangerschaft -- 10 Jahre Erfahrung in Oesterreich. Mitt Oester Ges Tropenmed Parasitol 1986;8:105-13.

33. Thalhammer O, Heller-Szollosy E. Erfahrungen mit routinemassigem Toxoplasmose-Screening bei Schwangeren zwecks Verhutung angeborener Toxoplasmose: eine prospektive Untersuchung. Wien Klin Wochenschr 1979;91:20-25.[Medline] 34. Flamm H, Aspock H. Die Toxoplasmose-Uberwachung der Schwangerschaft in Osterreich -- Ergebnisse und Probleme. Padiatr Grenzgeb 1981;20:27-34. [Medline] 35. Joynson DHM, Payne R. Screening for toxoplasma in pregnancy. Lancet 1988;2:795-796. 36. Prompeler HJ, Vogt A, Petersen EE. Toxoplasmose-Diagnostik in der Schwangerschaft. Gebursthilfe Frauenheilkd 1989;49:642-8. 37. Jeannel D, Costagliola D, Niel G, Hubert B, Danis M. What is known about the prevention of congenital toxoplasmosis? Lancet 1990;336:359-361.[Medline] 38. Desmonts G. Preventing congenital toxoplasmosis. Lancet 1990;336:10171018.[Medline] 39. Friese K, Beichert M, Hof H, et al. Untersuchung zur Haufigkeit konnataler Infektionen. Gebursthilfe Frauenheilkd 1991;51:890-6. 40. Toxoplasmose. In: Enders G. Infektionen und Impfungen in der Schwangerschaft. Munich, Germany: Urban & Schwarzenberg, 1988:143-61. 41. Hall SM. Congenital toxoplasmosis. BMJ 1992;305:291-297.[Medline] 42. Foulon W, Naessens A, Lauwers S, De Meuter F, Amy JJ. Impact of primary prevention on the incidence of toxoplasmosis during pregnancy. Obstet Gynecol 1988;72:363-366.[Medline]